Original Article NFATc4 and ATF3 Negatively Regulate Adiponectin Gene Expression in 3T3-L1 Adipocytes Hyun Bae Kim, Minyoung Kong, Tae Min Kim, Young Ho Suh, Won-Ho Kim, Joo Hyun Lim, Ji Hyun Song, and Myeong Ho Jung

Expression of adiponectin decreases with obesity and in- (TNF)-␣ (2). Adiponectin in particular possesses insulin- sulin resistance. At present, the mechanisms responsible sensitizing, antiatherogenic, anti-inflammatory, and antian- for negatively regulating adiponectin expression in adipo- giogenic properties (3,4). In humans and animals with cytes are poorly understood. In this investigation, we insulin resistance, obesity, or type 2 diabetes, serum levels -analyzed the effects of 5؅ serial deletion constructs on the of adiponectin are reduced (5–8). Although the physiolog murine adiponectin promoter. Here, we identified the re- ical effects of adiponectin have been investigated inten- pressor region located between ؊472 and ؊313 bp of the promoter. Removal of the putative nuclear factor of acti- sively, regulatory mechanisms for adiponectin gene vated T-cells (NFATs) binding site increased the promoter expression are poorly understood. Several insulin resis- activity, and overexpression of NFATc4 reduced the pro- tance–inducing factors such as TNF-␣ (9), interleukin moter activity. Treatment with the calcium ionophore (IL)-6 (10), and ␤-adrenergic agonists (11) have been A23187, an activator of NFAT, reduced mRNA as well as shown to reduce adiponectin expression. Conversely, in- promoter activity. The binding of NFATc4 to the promoter creases in adiponectin expression have been reported was associated with increased recruitment of histone during adipocyte differentiation and transcriptional fac- deacetylase 1 and reduced acetylation of histone H3 at the tors implicated in adipogenesis, including peroxisome promoter site. In addition, binding of activating transcrip- ␥ tion factor 3 (ATF3) to the putative activator protein-1 proliferator–activated receptor- (12), CCAAT/enhancer- site located adjacent to the NFAT binding site also re- binding protein (C/EBP)␣ (13,14), and sterol regulatory pressed the promoter activity. Treatment with thapsigar- element–binding protein-1c (15), have been shown to gin, an inducer of ATF3, reduced both mRNA and promoter upregulate adiponectin gene expression. However, little is activity. Importantly, the binding activities of NFATc4 and known about negative transcriptional control of adiponec- ATF3, increased significantly in white adipose tissues of tin gene expression. ob/ob and db/db mice compared with controls. Taken to- Nuclear factor of activated T-cell (NFAT) is a family of gether, this study demonstrates for the first time that transcription factors originally identified as important NFATc4 and ATF3 function as negative regulators of adi- mediators in cytokine gene expression during the immune ponectin gene expression, which may play critical roles in downregulating adiponectin expression in obesity and type response (16). However, recent evidence has demon- 2 diabetes. Diabetes 55:1342–1352, 2006 strated that NFAT is expressed in different cell types and regulates diverse cellular functions such as adipocyte differentiation (17), cardiac hypertrophy (18), neuronal development (19), and angiogenesis (20). Four members dipose tissues are known to store triglycerides of the NFAT family share significant sequence and func- and release free fatty acid/glycerol in response tional similarity. Calcium-sensitive NFATc1, -c2, and -c3 to changing energy demands (1). Additionally are tightly restricted to the immune system, whereas A however, adipose tissues also regulate energy NFATc4 is fairly ubiquitous and regulates cardiac hyper- homeostasis by secreting biologically active adipocyto- trophy and hippocampal neuronal signaling (21). NFAT kines, such as adiponectin, adipsin, leptin, plasminogen proteins are primarily phosphorylated and found in the activator inhibitor-1, resistin, and tumor necrosis factor cytoplasm of resting cells. Increased intracellular calcium levels activate calcineurin, a serine-threonine phosphatase that dephosphorylates NFAT, which then translocates into From the Division of Metabolic Diseases, Center for Biomedical Science, National Institute of Health, Seoul, Republic of Korea. the nucleus. Inside the nucleus, NFAT bind to a purine-rich Address correspondence and reprint requests to Myeong Ho Jung, Division core motif, (A/T)GGAAA, and regulate transcription of of Metabolic Diseases, Center for Biomedical Science, National Institute of NFAT-dependent genes (21). Cyclosporin A (CsA) and Health, 5 Nokbun-dong, Eunpyung-gu, Seoul, 122-701, Korea. E-mail: [email protected]. FK506 block calcineurin phosphatase activity and inhibit Received for publication 18 November 2005 and accepted in revised form 8 NFAT activation by preventing nuclear NFAT transloca- February 2006. tion (22). AP-1, activator protein-1; ATF3, activating transcription factor 3; C/EBP, CCAAT/enhancer-binding protein; ChIP, chromatin immunoprecipitation; Activating transcription factor 3 (ATF3) is a member of CsA, cyclosporin A; EMSA, electrophoretic mobility shift assay; HDAC, the ATF/cAMP responsive element–binding family of pro- histone deacetylase; HEK, human embryonic kidney; IL, interleukin; NFAT, teins that acts as a stress-inducible transcriptional repres- nuclear factor of activated T-cell; TNF, tumor necrosis factor; TSA, trichos- sor (23). ATF3 has been shown to be induced in cells tatin A. DOI: 10.2337/db05-1507 exposed to a variety of physiological and pathological © 2006 by the American Diabetes Association. stimuli, including carbon tetrachloride exposure (24), anti- The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance cancer drugs (25), proteasome inhibitors (26), genotoxic with 18 U.S.C. Section 1734 solely to indicate this fact. agents (27), homocysteine (28) and ischemia-reperfusion

1342 DIABETES, VOL. 55, MAY 2006 H.B. KIM AND ASSOCIATES

FIG. 1. Location of repressor element between ؊472 and ؊313 bp of the mouse adiponectin promoter. A: Constructs containing various lengths of serial deletions of the 5؅ flanking region of the mouse adiponectin gene were generated in pGL3 basic reporter plasmid. Luciferase activity is presented relative to the observable highest activity from the p(؊312)/Luc reporter. B: Schematic representation of mouse adiponectin promoter ؊472/؊313). The putative binding sites for NFAT, GKLF, and C/EBP are boxed. C: Deletion of putative NFAT site increases adiponectin) .promoter activity. The data are presented as means ؎ SD of five independent experiments

(29). ATF3 is also induced in response to endoplasmic (Promega, Madison, WI). Deletions in the 5Ј flanking regions of the adiponec- reticulum stress or amino acid deprivation (30). Activated tin promoter were constructed by PCR using a pairwise combination of the Ј Ј ATF3 can either homodimerize and repress transcription following sense primers: 5 -GGT ACC GAG GAT AAT TTT CAT TGC AC-3 [for P(Ϫ312)/Luc], 5Ј-GGT ACC CTC TTC ATT CTT ACT GAA AT-3Ј [for P(Ϫ472)/ of various promoters with ATF sites (31) or heterodimer- Luc], 5Ј-GGT ACC TTT GGC TGC ATG CAT ATT TG-3Ј [for P(Ϫ632)/Luc], ize with bZip proteins, c-jun, Jun B, ATF2, or gadd153/ 5Ј-GGT ACC ATG GTT CTC CAA TGT CAA GG-3Ј [for P(Ϫ860)/Luc], and CHOP10 (C/EBP homologous protein) and activate 5Ј-GGT ACC CCT AAC GTG ATT TCT CTA GA-3Ј [for P(Ϫ1,400)/Luc], with the transcription of target genes (32). antisense primer 5Ј-AGA TCT CTT TTG GTG TCG TCA GAT CC-3Ј. To understand mechanisms involved in regulating adi- Deletion-mutant plasmids lacking putative transcription factor binding ponectin gene expression, we isolated the mouse adi- sites and site-mutated plasmids were constructed in the pGL3 adiponectin ponectin promoter and analyzed the activities of luciferase plasmid by the two-step PCR method. Expression plasmids encod- Ј ing mouse NFATc1, c2, c3, or c4 were generous gifts from Dr. G. R. Crabtree promoters with 5 serial deletions. Here, we demonstrate (Department of Molecular Pharmacology, Albert Einstein College of Medi- that NFATc4 and ATF3 negatively regulate adiponectin cine). The ATF3 expression vector was kindly provided from Dr. T. Hai gene expression. The binding activities of both NFATc4 (Department of Molecular and Cellular Biochemistry, Ohio State University, and ATF3 from the nuclear extracts of white adipose Columbus, OH). tissue from ob/ob and db/db mice increase, inconsistent Transient transfection and luciferase activity. Fully differentiated 3T3-L1 with reductions in adiponectin mRNA. These results sug- adipocytes were grown in six-well plates and subjected to transient transfec- gest that NFATc4 and ATF3 play crucial roles in repres- tion using Lipofectamine reagent (Invitrogen) according to the manufacturer’s sion of adiponectin expression in obesity and type 2 instructions. After transfection for 24 h, cells were lysed in lysis buffer (Promega, Madison, WI), and luciferase activity was measured using the diabetes. Luciferase Assay System (Promega). Electrophoretic mobility shift assay. Probes corresponding to each pro- RESEARCH DESIGN AND METHODS moter binding region were generated, and electrophoretic mobility shift Cell culture and treatments. 3T3-L1 preadipocytes and human embryonic assays (EMSAs) were performed as previously described (13). kidney (HEK) 293 cell lines (ATCC, Manassas, VA) were cultured in Dulbec- RT-PCR. Total RNA was extracted using TRIzol reagent (Invitrogen) and was co’s modified Eagle’s medium with high glucose (Invitrogen, Carlsbad, CA) subjected to reverse transcription using reverse transcriptase (Promega) at supplemented with 10% (vol/vol) FCS (GibcoBRL, Gaithersburg, MD.). 3T3-L1 42°C for 1 h, and the resulting cDNA was amplified by PCR using gene-specific preadipocytes were differentiated as described previously (15). To investigate primers. the effects of stimulation on the adiponectin promoter, transfected 3T3-L1 Preparation of nuclear extracts. Nuclear extracts from 3T3-L1 adipocytes adipocytes were treated with 2 ␮mol/l of A23187 (Calbiochem, San Diego, were prepared as described by Crabtree (16). Nuclear extracts from mouse CA), 0.3 ␮mol/l of trichostatin A (TSA) (Biomol, Plymouth Meeting, PA), or 0.3 epididymal adipose tissues were performed as previously described (13). ␮mol/l of thapsigargin (Sigma, St Louis, MO) alone or in combination and 10 Western blot analysis. Total proteins were extracted by using PRO-PREP ng/ml of TNF-␣ (Calbiochem). For the determination of adiponectin mRNA reagent (Intron Biotechnology, Sungnam, Korea), and the immune complexes levels and assay of the chromatin immunoprecipitation (ChIP), the com- were identified using the enhanced chemiluminescence detection system pounds were added to fully differentiated 3T3-L1 cells for the indicated times. (Amersham Biosciences, Uppsala, Sweden). Plasmid constructs. The mouse adiponectin promoter region spanning ChIP. The ChIP assays were performed as described previously (33). Fully Ϫ1,500 to ϩ50 bp was amplified by PCR with mouse genomic DNA and differentiated 3T3-L1 cells were stimulated with the indicated compounds or inserted into KpnI/BglII restriction sites of the pGL3 basic luciferase reporter left unstimulated. Immunoprecipated DNA was amplified by PCR using

DIABETES, VOL. 55, MAY 2006 1343 NEGATIVE REGULATION OF ADIPONECTIN GENE

FIG. 2. NFATc4 downregulates transcription of adiponectin promoter dependent on putative NFAT binding site. A: NFATc4 represses adiponectin promoter activity in 3T3-L1 cells. The p(؊472)/Luc or p(؊860)/Luc reporter plasmids were transfected into 3T3-L1 adipocytes with the indicated ,NFAT expression vectors. B: NFATc4-mediated repression is dependent on putative NFAT binding site. The p(؊860)/Luc, p(؊472)/Luc p(؊312)/Luc reporters (left panel) and the mutant reporter lacking the NFAT binding site (right panel) were transfected into 3T3-L1 adipocytes .along with the NFATc4 expression vector. The data are presented as means ؎ SD of five independent experiments primers specific for the adiponectin promoter, sense (Ϫ432 to Ϫ413), 5Ј-GCT adiponectin promoter. A search of the TRANSFAC tran- TCA CAT TTA GTT ACA AA-3Ј and antisense (Ϫ292 to Ϫ273), 5Ј-ATA ATT scription factor binding database for a 160-bp consensus CAG CAT GTT TCT GA-3Ј or for exon 3 region, sense, 5Ј-GTCTCCACGACTCT Ј Ј Ј sequence revealed several potential binding sites for NFAT TACATG-3 and antisense, 5 -ACATTCATACACTCAGCCTG-3 . Ϫ Ϫ Ϫ Experimental animals. Obese (ob/ob) and diabetic (db/db) mice and their ( 363 to 344), gut-enriched kruppel-like factor ( 409 to age-matched control lean mice (C57BL/6J, 10 weeks) were purchased from Ϫ395), and C/EBP (Ϫ462 to Ϫ451) transcription factors, The Jackson Laboratory (Bar Harbor, ME). After overnight fasting, the mice all located within the Ϫ472 and Ϫ313 region (Fig. 1B). To were killed and the epididymal fat tissues collected and stored at Ϫ80°C for evaluate the functional significance of the putative repres- mRNA and nuclear protein isolation. sor binding sites, deletion-mutant constructs lacking bind- ing sites for the transcription factors were introduced into RESULTS the p(Ϫ472)/Luc reporter. As shown in Fig. 1C, the NFAT Location of repressor element between ؊472 to ؊313 site-deleted mutant increased promoter activity compared bp of the mouse adiponectin promoter. To dissect the with the wild type, whereas the gut-enriched kruppel-like regulatory DNA region involved in regulating adiponectin factor and C/EBP-deleted mutants produced minimal ef- gene expression, mouse adiponectin promoters with 5Ј fects on the promoter activity, indicating that NFAT may serial deletions were inserted into the pGL3 basic lucif- be involved in repressing adiponectin promoter activity. erase plasmid and transiently transfected into 3T3-L1 NFATc4 represses adiponectin promoter activity de- adipocytes. Analysis of the luciferase activities of trans- pendent on the putative NFAT binding site. To exam- fected cells revealed that the proximal Ϫ312-bp region ine whether NFAT is directly involved in repressing (Ϫ312 to ϩ39) of the adiponectin promoter conferred adiponectin expression, NFAT expression vectors were maximal transcriptional activity in 3T3-L1 cells (Fig. 1A). transfected along with p(Ϫ472)/Luc or p(Ϫ860)/Luc re- However, promoter activity was reduced by Ͼ70% with porters. As shown in Fig. 2A, exogenous overexpression of plasmid containing the Ϫ472-bp (Ϫ472 to ϩ39) region of NFATc4 significantly reduced the promoter activity of the adiponectin promoter (Fig. 1A). This suggests that the both p(Ϫ472)/Luc and p(Ϫ860)/Luc by 50% compared with presence of a repressor binding site was situated between the control vector. In contrast, the other NFAT family Ϫ472 and Ϫ313 bp of the 5Ј flanking region of the members, NFATc1, c2, and c3 all failed to reduce pro-

1344 DIABETES, VOL. 55, MAY 2006 H.B. KIM AND ASSOCIATES

FIG. 3. NFATc4 binds to the adiponectin promoter in vitro. A: Sequence comparison of putative NFAT binding site in the adiponectin promoter with the canonical NFAT binding site located on the IL-2 gene. The NFAT core binding site is boxed. Arrow indicates mutated site. B: EMSAs were performed on the NFAT site (؊363 to ؊344 bp) using nuclear extracts from 3T3-L1 adipocytes (left panel), murine white adipose tissue (middle panel), and NFATc4-overexpressing HEK293 cells (right panel). For oligonucleotide competition experiments, a 100-fold excess of oligonucle- otides were used (lane 1: no competition; lane 2: adiponectin wild NFAT oligonucleotides; lane 3: IL-2 NFAT oligonucleotides; lane 4: adiponectin mutant NFAT oligonucleotides; lane5: C/EBP oligonucleotides. The protein-DNA complex was supershifted by NFATc4 antibody (lane 6) but not by the nonspecific antibody (lane 7). moter activity, demonstrating that NFATc4 is likely re- panel) and mouse white adipose tissue (Fig. 3B, middle sponsible for repression of adiponectin expression in the panel), binding complexes were observed with the labeled natural promoter context. Next, we determined whether 20-bp oligonucleotides. A 100-fold molar excess of unla- NFATc4-dependent repression is mediated through the beled oligonucleotides for homologous adiponectin and repressor region containing the putative NFAT binding IL-2 NFAT nearly abolished the bands, whereas unlabeled site. As shown in Fig. 2B (left panel), the promoter mutant oligonucleotides bore no significant effect on bind- activities of both p(Ϫ860)/Luc and p(Ϫ472)/Luc reporters ing (Fig. 3B). To ascertain whether the complex consisted were reduced by NFATc4, while the p(Ϫ312)/Luc reporter, of bound NFATc4, a supershift assay was conducted with which does not contain the repressor region, was unaf- the antibody against NFATc4. Results show that anti- fected. To further confirm the role of the putative NFAT NFATc4 shifted the binding complex. In contrast, no binding site, an NFAT binding site–deleted reporter was supershifted complex was detected with anti-STAT-1 (Fig. constructed in p(Ϫ472)/Luc. As shown in Fig. 2B (right 3B). Nuclear extracts from HEK293 cells overexpressing panel), NFATc4 repressed the promoter activity of wild- NFATc4 were also used to examine specific binding of type p(Ϫ472)/Luc but did not affect the NFAT site–deleted NFATc4 to the binding site. The overexpression of reporter. Taken together, these results demonstrate that NFATc4 was confirmed by Western blot (data not shown). NFATc4 represses the promoter activity of adiponectin As shown in Fig. 3B (right panel), overexpressing NFATc4 genes, dependent upon the putative NFAT binding site formed complexes with the labeled oligonucleotides and (Ϫ363 to Ϫ344) of the mouse adiponectin promoter. competed against oligonucleotides for homologous adi- NFATc4 binds to putative NFAT binding site on the ponectin and IL-2 NFAT. Moreover, the NFATc4 antibody adiponectin promoter. To determine whether NFATc4 successfully supershifted the complex. All together, these interacts with the NFAT-binding site, EMSAs were per- data demonstrate that NFATc4 directly binds to the NFAT formed using nuclear extracts from fully differentiated binding site located between Ϫ363 and Ϫ344 of the 3T3-L1 cells and mouse white adipose tissue. For this adiponectin promoter. experiment, a 20-bp wild-type oligonucleotide (Ϫ363 to NFATc4 activation reduces mRNA expression and Ϫ344) was designed that cover the putative NFAT binding promoter activity of adiponectin. To address whether site containing the core sequence of the NFAT response NFATc4 activation downregulates adiponectin expression, element (GGAAA) (Fig. 3A). As shown in Fig. 3B, using fully differentiated 3T3-L1 cells were treated with A23187 nuclear extracts from both 3T3-L1 cells (Fig. 3B, left that activates NFAT, and adiponectin mRNA levels were

DIABETES, VOL. 55, MAY 2006 1345 NEGATIVE REGULATION OF ADIPONECTIN GENE

FIG. 4. NFATc4 activation reduces adiponectin promoter activity and mRNA expression. A: A23187 reduces mRNA expression of adiponec- tin. 3T3-L1 adipocytes were treated with A23187 in the absence or presence of CsA for the indicated times. Total RNA were subjected to RT-PCR using adiponectin-specific primers. B: A23187 represses adi- ponectin promoter activity. 3T3-L1 adipocytes transfected with p(؊860)/Luc reporter were left unstimulated or stimulated with A23187 for 24 h in the absence or presence of CsA. C: A23187 increases NFATc4-dependent repression of adiponectin promoter. 3T3-L1 adipo- -cytes cotransfected with p(؊860)/Luc reporter and NFATc4 expres sion vector were left unstimulated or stimulated with A23187 for 24 h. D: Downregulation of adiponectin promoter activity by A23187 is dependent on the NFAT binding site. A site-directed mutatation :GGGAAA3CTTAAA) was introduced to the p(؊860)/Luc reporter. E) A23187 increases binding activity of NFATc4 on the adiponectin promoter. EMSA was performed using nuclear extracts from A23187- treated 3T3-L1 adipocytes in the absence or presence of CsA. measured by RT-PCR. As shown in Fig. 4A, A23187 signif- ponectin gene is mediated through increased NFATc4 icantly reduced mRNA. However, treatment with CsA, an binding to the adiponectin promoter, EMSA was per- inhibitor of NFAT, reversed the repressive effect of formed using nuclear extracts from A23187-treated 3T3-L1 A23187. We next determined whether reductions in adi- adipocytes. As shown in Fig. 4E, the NFATc4 binding ponectin mRNA by A23187 is correlated with decreased complex was strongly induced by A23187 treatment and transcriptional activity mediated by the adiponectin pro- pretreatment with CsA abolished induction of the binding moter. Consistent with decreased mRNA, A23187 also complexes. repressed promoter activity of p(Ϫ860)/Luc, which was NFATc4 interacts with adiponectin promoter in vivo, reversed by CsA treatment (Fig. 4B). Moreover, combining and NFATc4-mediated repression is associated with overexpression of NFATc4 and A23187 treatment further recruitment of histone deacetylase activity. To exam- reduced the promoter activity of p(Ϫ860)/Luc (Fig. 4C). ine binding of NFAT to the adiponectin promoter in vivo, However, constructs containing the mutant NFAT binding ChIP was performed in 3T3-L1 adipocytes stimulated with site diminished the transcriptional repression of the adi- A23187 and unstimulated cells. The immunoprecipitated ponectin promoter from A23187 stimulation (Fig. 4D). chromatins were amplified by PCR using adiponectin These data suggest that NFATc4 activation represses promoter-specific primers (Fig. 5A). As shown in Fig. 5B, promoter activity dependent on the NFAT binding site. To the 160-bp PCR product was compared with the NFATc4 identify whether A23187-mediated repression of the adi- antibody, whereas no bands were detected with primers in

1346 DIABETES, VOL. 55, MAY 2006 H.B. KIM AND ASSOCIATES

FIG. 5. NFATc4 binds to the adiponectin promoter in vivo, whose binding reduces histone H3 specific-acetylation. A: Diagram showing the oligonucleotides used to amplify the adiponectin promoter in the ChIP assay. B: ChIP assay was performed with or without antibody against NFATc4. 3T3-L1 adipocytes were incubated with A23187 in the absence or presence of CsA. C: A23187 decreases acetylation of histone H3. 3T3-L1 adipocytes were treated with A23187 in the absence or presence of CsA. ChIP assay was performed with or without antibody against acetyl H3, acetyl H4, and HDAC1. D: TSA reverses A23187-dependent repression of adiponectin promoter. 3T3-L1 adipocytes transfected with p(؊860)/Luc reporter were left unstimulated or stimulated with A23187 for 24 h in the absence or presence of TSA. Luciferase activity is presented relative .to unstimulation. The data are represented as means ؎ SD of five independent experiments exon3, which is located 10 kb downstream of NFAT negatively regulates adiponectin promoter through histone binding region, indicating that NFATc4 formed a specific H3-specific deacetylation. complex with the adiponectin promoter in vivo. The ATF3 also represses transcriptional activity of the intensity of the PCR product was increased with stimula- adiponectin gene. It has been demonstrated that acti- tion of A23187 compared with nonstimulation. Moreover, vated NFAT interacts with other transcription factors such treatment with CsA reversed the stimulated intensity (Fig. as MEF2, GATA2, GATA4, or, more frequently, with acti- 5B), indicating that NFAT activation increased binding of vator protein-1 (AP-1) at composite DNA elements to NFATc4 to the promoter in vivo. regulate gene expression (18,35,36). In the adiponectin Previously, a published report found that NFATc2 re- promoter, a putative AP-1 binding site (TGACTCTC, Ϫ376/ pressed CDK4 promoter activity by reducing acetylation of Ϫ369) was found 15 bp apart from the core NFAT binding histone H3 through recruitment of a histone deacetylase sequence. To test the functional role of this site, a deletion (HDAC) family member to the CDK promoter (34). There- mutant was generated using the p(Ϫ472)/Luc reporter. As fore, we investigated whether NFATc4-dependent repres- shown in Fig. 6A, removal of the putative AP-1 binding site sion of adiponectin promoter is also regulated by HDAC increased promoter activity equivalent to that of the NFAT activity. To this end, we utilized ChIP analysis to examine binding site-deleted mutant, suggesting that the putative the status of histone H3 and H4 acetylation of the adi- AP-1 binding site may also be involved in repressing the ponectin promoter when treated with A23187. As shown in adiponectin promoter. Next, we investigated whether AP-1 Fig. 5C, A23187 decreased histone H3 acetylation, but left could directly regulate the adiponectin promoter. When acetylation of histone H4 unchanged. However, the addi- the p(Ϫ860)/Luc reporter was cotransfected with the c-jun tion of CsA reversed the histone H3 acetylation to levels expression vector into 3T3-L1 adipocyte cells, it did not comparable to unstimulated cells. To characterize the affect promoter activity (data not shown). Since it had involvement of HDAC in NFATc4-mediated decrease in been reported that ATF3, a stress-responsive transcription histone H3 acetylation, recruitment of HDAC1 to the repressor, can effectively bind to the AP-1 site and repress adiponectin promoter was investigated in A23187-stimu- transcriptional activity (30), we examined whether ATF3 lated cells by ChIP analysis. As shown in Fig. 5C, A23187 represses adiponectin gene expression. As shown in Fig. treatment increased recruitment of HDAC1, which was 6B, expression of ATF3 repressed the promoter activity of prevented by treatment with CsA. Moreover, to further both the p(Ϫ860)/Luc and p(Ϫ472)/Luc reporters, whereas confirm these findings, the effect of TSA, an irreversible it did not repress the promoter activity of the p(Ϫ312)/Luc HDAC inhibitor, on A23187-mediated repression of pro- reporter missing the AP-1 binding site. These findings moter activity was examined. As shown in Fig. 5D, treat- demonstrate that ATF3 downregulates the adiponectin ment with TSA potently reversed repression of the promoter. Thapsigargin is a known activator of ATF3 and promoter by A23187. These results suggest that NFATc4 inducer of endoplasmic reticulum stress. Therefore, we

DIABETES, VOL. 55, MAY 2006 1347 NEGATIVE REGULATION OF ADIPONECTIN GENE

FIG. 6. ATF3 downregulates transcription of the adiponectin pro- moter. A: Schematic representation of luciferase reporters used in transfection assay. Deletion of putative AP-1 binding site increases adiponectin promoter activity. B: Expression of ATF3 represses adi- ponectin promoter activity. The p(؊312)/Luc, p(؊472)/Luc, and p(؊860)/Luc were transfected into 3T3-L1 adipocytes along with ATF3 expression vectors. C: Thapsigargin reduces mRNA expression of adiponectin. 3T3-L1 adipocytes were treated with thapsigargin as indicated. Total RNA were subjected to RT-PCR using gene-specific primers. D: Thapsigargin represses adiponectin promoter activity. -3T3-L1 adipocytes transfected with p(؊860)/Luc were left unstimu lated or stimulated with thapsigargin for 24 h. E: ATF3 represses /(adiponectin promoter activity with NFATc4 additively. The p(؊860 Luc was transfected into 3T3-L1 adipocytes with ATF3 or NFATc4 and both expression vectors. used thapsigargin to investigate whether ATF3 activation antibody. As shown in Fig. 7B, anti-ATF3 inhibited binding represses transcription of the adiponectin gene. As shown of complexes formed in the nuclear extracts, whereas no in Fig. 6C, treatment of fully differentiated 3T3-L1 cells inhibition was detected with anti-STAT1 (Fig. 7B), indicat- with thapsigargin significantly reduced adiponectin mRNA ing that ATF3 specifically binds to the putative AP-1 with a concomitant increase in ATF3 levels, as well as binding site of adiponectin promoter. To examine binding efficiently repressing the promoter activity of the p(Ϫ860)/ of ATF3 to the adiponectin promoter in vivo, ChIP was Luc reporter (Fig. 6D). In combination, NFATc4 and ATF3 performed. The 160-bp PCR product was produced from additively repressed promoter activity by Ͼ80% (Fig. 6E) the immunoprecipitated chromatins (Fig. 7C, left panel), compared with ATF3 or NFATc4 expression alone. and the intensity of the PCR product was increased with ATF3 binds to the adiponectin promoter in vitro and thapsigargin treatment compared with unstimulated cells in vivo. To demonstrate whether ATF3 binds to the (Fig. 7C, left panel). In agreement with this result, thapsi- putative AP-1 site of the adiponectin promoter, we per- gargin treatment increased ATF3 binding activity on adi- formed EMSAs using nuclear extracts from fully differen- ponectin promoter (Fig. 7C, right panel). Therefore, these tiated 3T3-L1 cells, mouse white adipose tissue, and results suggest that reduced adiponectin expression by ATF3-overexpressing HEK293 cells. Using the labeled oli- thapsigargin may be mediated through increased ATF3 gonucleotide covering the Ϫ383/Ϫ364-bp region (Fig. 7A), binding to the adiponectin promoter. Furthermore, since binding complexes were observed in the nuclear extracts recent studies have reported that TNF-␣, whose expres- (Fig. 7B). Unlabeled ATF/cAMP response element consen- sion increased in obesity animal models, induces ATF3 in sus oligonucleotides, as well as homologous oligonucleo- vascular endothelial cells (37), we investigated whether tides, nearly eliminated the retarded bands, but the TNF-␣ also induces ATF3 expression in 3T3-L1 adipocytes. unlabeled mutant oligonucleotides exerted no significant As shown in Fig. 7D, treatment of TNF-␣ significantly effect on binding. To further confirm ATF3 binding, the increased ATF3 expression (left panel) and the binding antibody supershift assay was conducted with the ATF3 activity on adiponectin promoter (right panel), strongly

1348 DIABETES, VOL. 55, MAY 2006 H.B. KIM AND ASSOCIATES

FIG. 7. ATF3 binds to the adiponectin promoter in vitro and in vivo. A: Sequence comparison of the putative AP-1 site in the adiponectin promoter with the canonical AP-1 binding site. The AP-1 consensus sequence is boxed. Arrow indicates mutated site. B: EMSAs were performed on the putative AP-1 site using nuclear extracts from 3T3-L1 adipocytes (left panel), murine white adipose tissue (middle panel), and ATF3- overexpressing HEK293 cells (right panel). For oligonucleotide competition experiments, a 100-fold excess was used (lane 1: no competition; lane 2: adiponectin wild AP-1 oligonucleotides; lane 3: consensus ATF/cAMP response element oligonucleotides; lane 4: adiponectin mutant AP-1 oligonucleotides; lane 5: C/EBP oligonucleotides). The protein-DNA complex was diminished by anti ATF-3 (lane 6), but not by the nonspecific antibody (lane 7). C: ChIP assay was performed with or without antibody against ATF3 (left panel). The immunoprecitated DNA fragments from thapsigargin-treated 3T3-L1 adipocytes were amplified by PCR with primers used in Fig. 5. Thapsigargin increases the binding activity of ATF3 on adiponectin promoter (right panel). D: TNF-␣ increases the expression of ATF3 (left panel) and the binding activity (right panel). suggesting that suppression of adiponectin in obesity by mice by EMSAs. As shown in Fig. 8A, the binding activities TNF-␣ may be partly mediated through ATF3. of both NFATc4 and ATF-3 increased significantly in ob/ob Increase in binding activities of both NFATc4 and and db/db mice compared with control mice, whereas Sp1 ATF3 to the adiponectin promoter from the nuclear used as a control and binding activity was unchanged in extracts of the white adipocytes of ob/ob and db/db nuclear extracts, suggesting that both NFATc4 and ATF3 mice. Since adiponectin expression is significantly re- may play critical roles in repression of adiponectin expres- duced in the white adipose tissue of obese and diabetic sion in obese and diabetic subjects. Moreover, when animal models, we examined the involvement of NFATc4 expression of NFATc4 or ATF3 was examined from white and ATF3 in the repression of adiponectin expression due adipose tissue of ob/ob or db/db mice, ATF3 was drastically to obesity and type 2 diabetes. To this end, we examined upregulated in both types of mice consistent with reduc- the binding activities of NFATc4 and ATF3 in nuclear tions in adiponectin mRNA, whereas NFATc4 was not extracts from the white adipose tissue of ob/ob and db/db significantly changed (Fig. 8B). Therefore, increased ATF3

DIABETES, VOL. 55, MAY 2006 1349 NEGATIVE REGULATION OF ADIPONECTIN GENE

FIG. 8. The binding activities of NFATc4 and ATF3 increase in nuclear extracts from the white adipose tissue of ob/ob and db/db mice. A: Nuclear extracts were prepared from the white adipose tissues of ob/ob and db/db mice and analyzed by EMSA using probes for NFATc4 (left panel), ATF3 (middle panel), and SP-1 as a control (right panel). B: The mRNA expression of ATF3 increases in white adipose tissue of ob/ob and db/db mice, concomitant with an increase in endoplasmic reticulum stress. Total RNA were isolated from the white adipose tissues of ob/ob and db/db mice and subjected to RT-PCR using primers for ATF3, NFATc4, and adiponectin. C: Endoplasmic reticulum stress increases in white adipose tissue of ob/ob and db/db mice. Western blotting was performed for the detection of CHOP and eIF2␣ phosphorylation. binding activity in ob/ob and db/db mice may result from collagen (38). In this study, we demonstrate that NFATc4 elevated ATF3 expression, whereas NFATc4 may be acti- suppresses adiponectin gene expression in 3T3-L1 adipo- vated by calcineurin signaling in obesity and type 2 diabe- cytes dependent on an NFAT binding site (Ϫ363 to Ϫ344) tes. Since endoplasmic reticulum stress–inducible ATF3 in the mouse adiponectin promoter. To understand the was upregulated in ob/ob and db/db mice, we then exam- mechanism involved in negative regulation of adiponectin ined the expression of molecular indicators of endoplas- expression by NFATc4, we investigated chromatin remod- mic reticulum stress in the white adipose tissue of ob/ob eling–associated factors since chromatin structure remod- and db/db mice. As shown in Fig. 8C, expression of CHOP eling such as acetylation and deacetylation of histone and eIF2␣ phosphorylation increased in ob/ob and db/db plays a critical role in regulating gene expression. Both mice compared with controls, suggesting that induction of histone H3 and H4 acetylation allows transacting factors endoplasmic reticulum stress in obesity and type 2 diabe- to associate with cognate DNA binding sites and transcrip- tes may involve in downregulation of adiponectin expres- tion to occur (39,40). Conversely, HDAC reverses histone sion via ATF3. acetylation, resulting in a more compact chromatin that is transcriptionally repressive. Our study results show that DISCUSSION under conditions that maximize promoter repression by Even though several positive transcription factors includ- treatment with A23187, reductions in histone H3 acetyla- ing peroxisome proliferator–activated receptor-␥ (12), tion within the adiponectin promoter region were ob- C/EBP␣ (13,14), and ADD1/sterol regulatory element– served together with recruitment of HDAC1. This suggests binding protein-1c (15), have been identified as regulators that increased NFAT binding to the adiponectin promoter of adiponectin gene expression, there is little information reduces H3 acetylation through recruitment of HDAC1. about negative transcription factors, particularly under These events are inhibited by the addition of the NFAT conditions of obesity, insulin resistance, or type 2 diabe- inhibitor, CsA. Even though we did not investigate a direct tes. Here, we provide evidence that both NFATc4 and association between HDAC1 and NFATc4, repression of ATF3 negatively regulate adiponectin gene expression in the adiponectin gene expression by NFATc4 may be partly 3T3-L1 adipocytes, which may contribute to repression of mediated by reduced histone acetylation associated with adiponectin expression in obese and diabetic subjects. binding of NFATc4 in mature adipocytes. Several lines of evidence indicate that NFAT is involved Here, we also identified that ATF3 additionally functions in stimulating transcription of inducible genes in immune as a transcriptional repressor by binding to a putative AP-1 and nonimmune cells, but very recently, a regulatory role site adjacent to the NFAT binding site of the adiponectin for NFAT has been reported where NFATc2 suppresses promoter. It has been demonstrated that ATF3 ho- CDK4 gene expression through binding to a site immedi- modimers recognize both the ATF/cAMP response ele- ately downstream of the transcriptional start site (34) and ment site and the AP-1 site and plays a transcriptional expression of cartilage markers, type II, and type X repressor role. Therefore, ATF3 can repress expression of

1350 DIABETES, VOL. 55, MAY 2006 H.B. KIM AND ASSOCIATES itself (41), its inhibitor (gadd153/Chop10) (42), PEPCK REFERENCES (43), and the TNF-␣–induced E-selectin gene (44). In the 1. Spiegelman BM, Flier JS: Adipogenesis and obesity: rounding out the big present study, we verified that ATF3 binds to a putative picture. Cell 87:377–389, 1996 AP-1 site on the adiponectin promoter and that expression 2. Mohamed-Ali V, Pinkney JH, Coppack SW: Adipose tissue as an endocrine and paracrine organ. Int J Obes Relat Metab Disord 22:1145–1158, 1998 of ATF3 significantly reduces promoter activity. Impor- 3. Diez JJ, Iglesias P: The role of the novel adipocyte-derived hormone tantly, treatment with thapsigargin, an endoplasmic retic- adiponectin in human disease. Eur J Endocrinol 148:293–300, 2003 ulum stress inducer, reduces adiponectin gene expression 4. Chandran M, Phillips SA, Ciaraldi T, Henry RR: Adiponectin: more than just together with an increase in ATF3. Therefore, it is sug- another fat cell hormone? Diabetes Care 26:2442–2450, 2003 gested that endoplasmic reticulum stress is implicated in 5. Hotta K, Funahashi T, Arita Y, Takahashi M, Matsuda M, Okamoto Y, Iwahashi H, Kuriyama H, Ouchi N, Maeda K, Nishida M, Kihara S, Sakai N, downregulating adiponectin expression through induction Nakajima T, Hasegawa K, Muraguchi M, Ohmoto Y, Nakamura T, Ya- of ATF3. Studying the cooperation of NFATc4 and ATF3 in mashita S, Hanafusa T, Matsuzawa Y: Plasma concentrations of a novel, regulating adiponectin promoter, both NFATc4 and ATF3 adipose-specific protein, adiponectin, in type 2 diabetic patients. Arterio- appear to negatively regulate adiponectin gene expression scler Thromb Vasc Biol 20:1595–1599, 2000 independently. The coexistence of both NFATc4 and ATF3 6. Tsao TS, Lodish HF, Fruebis J: ACRP30, a new hormone controlling fat and results in additive repression of adiponectin expression. glucose metabolism. Eur J Pharmacol 440:213–221, 2002 7. Arita Y, Kihara S, Ouchi N, Takahashi M, Maeda K, Miyagawa J, Hotta K, The mechanism responsible for the negative regulation Shimomura I, Nakamura T, Miyaoka K, Kuriyama H, Nishida M, Yamashita of adiponectin in obesity and insulin resistance have not S, Okubo K, Matsubara K, Muraguchi M, Ohmoto Y, Funahashi T, Matsu- been determined in details so far. As described earlier, zawa Y: Paradoxical decrease of an adipose-specific protein, adiponectin, increased TNF-␣ and IL-6 may partially be responsible for in obesity. Biochem Biophys Res Commun 257:79–83, 1999 decreased adiponectin production in obese subjects 8. Statnick MA, Beavers LS, Conner LJ, Corominola H, Johnson D, Hammond CD, Rafaeloff-Phail R, Seng T, Suter TM, Sluka JP, Ravussin E, Gadski RA, (9,10). Based on our present findings that the binding Caro JF: Decreased expression of apM1 in omental and subcutaneous activities of both NFATc4 and ATF3 increases in the adipose tissue of humans with type 2 diabetes. Int J Exp Diabetes Res nuclear extracts of white adiopose tissue of ob/ob and 1:81–88, 2000 db/db mice and ATF3 expression also increases in white 9. Fasshauer M, Klein J, Neumann S, Eszlinger M, Paschke R: Hormonal adiopose tissue of ob/ob and db/db mice, we suggest that regulation of adiponectin gene expression in 3T3–L1 adipocytes. Biochem Biophys Res Commun 290:1084–1089, 2002 NFATc4 and ATF3 play crucial roles in reducing adiponec- 10. Fasshauer M, Kralisch S, Klier M, Lossner U, Bluher M, Klein J, Paschke R: tin gene expression in obesity and type 2 diabetes. More- Adiponectin gene expression and secretion is inhibited by interleukin-6 in over, TNF-␣ significantly increases ATF3 expression in 3T3-L1 adipocytes. Biochem Biophys Res Commun 301:1045–1050, 2003 3T3-L1 cells with reduction of adiponectin gene expres- 11. Fasshauer M, Klein J, Neumann S, Eszlinger M, Paschke R: Adiponectin gene expression is inhibited by beta-adrenergic stimulation via protein sion. Therefore, we also suggest that ATF3 may involve in kinase A in 3T3-L1 adipocytes. FEBS Lett 507:142–146, 2001 TNF-␣–mediated downregulation of adiponectin gene in 12. Iwaki M, Matsuda M, Maeda N, Funahashi T, Matsuzawa Y, Makishima M, the adipocyte under obesity states. Obesity causes endo- Shimomura I: Induction of adiponectin, a fat-derived antidiabetic and plasmic reticulum stress, which in turn leads to impaired antiatherogenic factor, by nuclear receptors. Diabetes 52:1655–1663, 2003 insulin resistance and type 2 diabetes through suppression 13. Park SK, Oh SY, Lee MY, Yoon S, Kim KS, Kim JW: CCAAT/enhancer binding protein and nuclear factor-Y regulate adiponectin gene expression of insulin receptor signaling by the hyperactivation of in adipose tissue. Diabetes 53:2757–2766, 2004 c-Jun NH2-terminal kinase and subsequent serine phos- 14. Qiao L, Maclean PS, Schaack J, Orlicky DJ, Darimont C, Pagliassotti M, phorylation of insulin receptor substrate-1 in the liver and Friedman JE, Shao J: C/EBP␣ regulates human adiponectin gene transcrip- adipocytes (45). Since our present study shows that ATF3 tion through an intronic enhancer. Diabetes 54:1744–1754, 2005 15. Seo JB, Moon HM, Noh MJ, Lee YS, Jeong HW, Yoo EJ, Kim WS, Park J, induced by endoplasmic reticulum stress represses adi- Youn BS, Kim JW, Park SD, Kim JB: Adipocyte determination- and ponectin expression, endoplasmic reticulum stress may be differentiation-dependent factor 1/sterol regulatory element-binding pro- responsible for repression of adiponectin gene expression tein 1c regulates mouse adiponectin expression. J Biol Chem 279:22108– due to obesity and type 2 diabetes via ATF3. Furthermore, 22117, 2004 this finding strongly supports another role of endoplasmic 16. Crabtree GR: Calcium, calcineurin, and the control of transcription. J Biol Chem 276:2313–2316, 2001 reticulum stress in causing insulin resistance and type 2 17. Ho IC, Kim JH, Rooney JW, Spiegelman BM, Glimcher LH: A potential role diabetes through downregulation of adiponectin expres- for the nuclear factor of activated T cells family of transcriptional sion in adipocyte. regulatory proteins in adipogenesis. Proc Natl Acad SciUSA95:15537– In summary, this is the first report showing that NFATc4 15541, 1998 18. Molkentin JD, Lu JR, Antos CL, Markham B, Richardson J, Robbins J, and ATF3 act as transcriptional repressors in regulating Grant SR, Olson EN: A calcineurin-dependent transcriptional pathway for adiponectin gene expression, providing a potential mech- cardiac hypertrophy. Cell 93:215–228, 1998 anism for downregulating adiponectin expression in obe- 19. Graef IA, Chen F, Crabtree GR: NFAT signaling in vertebrate development. sity and type 2 diabetes, and implicating an additional role Curr Opin Genet Dev 11:505–512, 2001 of endoplasmic reticulum stress in causing insulin resis- 20. Zaichuk TA, Shroff EH, Emmanuel R, Filleur S, Nelius T, Volpert OV: Nuclear factor of activated T cells balances angiogenesis activation and tance and type 2 diabetes. inhibition. J Exp Med 199:1513–1522, 2004 21. Hogan PG, Chen L, Nardone J, Rao A: Transcriptional regulation by calcium, calcineurin, and NFAT. Genes Dev 17:2205–2232, 2003 ACKNOWLEDGMENTS 22. Flanagan WM, Corthesy B, Bram RJ, Crabtree GR: Nuclear association of a T-cell transcription factor blocked by FK-506 and cyclosporin A. Nature This work was supported by an intramural grant from the 352:803–807, 1991 National Institute of Health, Korea (347-6111-211-207). 23. Hashimoto Y, Zhang C, Kawauchi J, Imoto I, Adachi MT, Inazawa J, We are grateful to Dr. G.R. Crabtree (Albert Einstein Amagasa T, Hai T, Kitajima S: An alternatively spliced isoform of tran- College) for providing us with NFAT expression vectors scriptional repressor ATF3 and its induction by stress stimuli. Nucleic Acid Res 30:2398–2406, 2002 and to Dr. T. Hai (Ohio State University) for his gift of the 24. Chen BP, Wolfgang CD, Hai T: Analysis of ATF3, a transcription factor ATF3 expression vector. We also thank Dr. Van-Anh induced by physiological stresses and modulated by gadd153/Chop10. Mol Nguyen for peer reviewing the manuscript. Cell Biol 16:1157–1168, 1996

DIABETES, VOL. 55, MAY 2006 1351 NEGATIVE REGULATION OF ADIPONECTIN GENE

25. Shtil AA, Mandlekar S, Yu R, Walter RJ, Hagen K, Tan TH, Roninson IB, 35. Blaeser F, Ho N, Prywes R, Chatila TA: Ca(2ϩ)-dependent gene expression Kong AN: Differential regulation of mitogen-activated protein kinases by mediated by MEF2 transcription factors. J Biol Chem 275:197–209, 2000 microtubule-binding agents in human breast cancer cells. Oncogene 18: 36. Musaro A, McCullagh KJ, Naya FJ, Olson EN, Rosenthal N: IGF-1 induces 377–384, 1999 skeletal myocyte hypertrophy through calcineurin in association with 26. Zimmermann J, Erdmann D, Lalande I, Grossenbacher R, Noorani M, Furst GATA-2 and NF-ATc1. Nature 400:581–585, 1999 P: Proteasome inhibitor induced gene expression profiles reveal overex- 37. Inoue K, Zama T, Kamimoto T, Aoki R, Ikeda Y, Kimura H, Hagiwara M: pression of transcriptional regulators ATF3, GADD153 and MAD1. Onco- TNFalpha-induced ATF3 expression is bidirectionally regulated by the JNK gene 19:2913–2920, 2000 and ERK pathways in vascular endothelial cells. Genes Cells 9:59–70, 2004 27. Amundson SA, Bittner M, Chen Y, Trent J, Meltzer P, Fornace AJ Jr: 38. Ranger AM, Gerstenfeld LC, Wang J, Kon T, Bae H, Gravallese EM, Fluorescent cDNA microarray hybridization reveals complexity and het- Glimcher MJ, Glimcher LH: The nuclear factor of activated T cells (NFAT) erogeneity of cellular genotoxic stress responses. Oncogene 18:3666–3672, transcription factor NFATp (NFATc2) is a repressor of chondrogenesis. J 1999 Exp Med 191:9–22, 2000 28. Cai Y, Zhang C, Nawa T, Aso T, Tanaka M, Oshiro S, Ichijo H, Kitajima S: 39. Wolffe AP, Guschin D: Review: chromatin structural features and targets Homocysteine-responsive ATF3 gene expression in human vascular endo- that regulate transcription. J Struct Biol 129:102–122, 2000 thelial cells: activation of c-Jun NH(2)-terminal kinase and promoter 40. Roth SY, Denu JM, Allis CD: Histone acetyltransferases. Annu Rev response element. Blood 96:2140–2148, 2000 Biochem 70:81–120, 2001 29. Yin T, Sandhu G, Wolfgang CD, Burrier A, Webb RL, Rigel DF, Hai T, 41. Wolfgang CD, Liang G, Okamoto Y, Allen AE, Hai T: Transcriptional Whelan J: Tissue-specific pattern of stress kinase activation in ischemic/ autorepression of the stress-inducible gene ATF3. J Biol Chem 275:16865– reperfused heart and kidney. J Biol Chem 272:19943–19950, 1997 16870, 2000 30. Jiang HY, Wek SA, McGrath BC, Lu D, Hai T, Harding HP, Wang X, Ron D, 42. Fawcett TW, Martindale JL, Guyton KZ, Hai T, Holbrook NJ: Complexes Cavener DR, Wek RC: Activating transcription factor 3 is integral to the eukaryotic initiation factor 2 kinase stress response. Mol Cell Biol 24:1365– containing activating transcription factor (ATF)/cAMP-responsive-ele- 1377, 2004 ment-binding protein (CREB) interact with the CCAAT/enhancer-binding 31. Chen BP, Liang G, Whelan J, Hai T: ATF3 and ATF3 delta Zip: transcrip- protein (C/EBP)-ATF composite site to regulate Gadd153 expression tional repression versus activation by alternatively spliced isoforms. J Biol during the stress response. Biochem J 339:135–141, 1999 Chem 269:15819–15826, 1994 43. Allen-Jennings AE, Hartman MG, Kociba GJ, Hai T: The roles of ATF3 in 32. Hsu JC, Bravo R, Taub R: Interactions among LRF-1, JunB, c-Jun, and liver dysfunction and the regulation of phosphoenolpyruvate carboxyki- c-Fos define a regulatory program in the G1 phase of liver regeneration. nase gene expression. J Biol Chem 277:20020–20025, 2002 Mol Cell Biol 12:4654–4665, 1992 44. Nawa T, Nawa MT, Cai Y, Zhang C, Uchimura I, Narumi S, Numano F, 33. Latasa MJ, Griffin MJ, Moon YS, Kang C, Sul HS: Occupancy and function Kitajima S: Repression of TNF-alpha-induced E-selectin expression by of the Ϫ150 sterol regulatory element and Ϫ65 E-box in nutritional PPAR activators: involvement of transcriptional repressor LRF-1/ATF3. regulation of the fatty acid synthase gene in living animals. Mol Cell Biol Biochem Biophys Res Commun 275:406–411, 2000 23:5896–5907, 2003 45. Ozcan U, Cao Q, Yilmaz E, Lee AH, Iwakoshi NN, Ozdelen E, Tuncman G, 34. Baksh S, Widlund HR, Frazer-Abel AA, Du J, Fosmire S, Fisher DE, Gorgun C, Glimcher LH, Hotamisligil GS: Endoplasmic reticulum stress DeCaprio JA, Modiano JF, Burakoff SJ: NFATc2-mediated repression of links obesity, insulin action, and type 2 diabetes. Science 306:457–461, cyclin-dependent kinase 4 expression. Mol Cell 10:1071–1081, 2002 2004

1352 DIABETES, VOL. 55, MAY 2006